Elasticsearch collector multi instance problem

[AlexAkulov]

Hello!
I have three instances of Elasticsearch on one host.

enabled = true
instances = instance1@localhost:9202,instance2@localhost:9204,instance3@localhost:9206
path_prefix = EDI.elasticsearch
logstash_mode = True
stats = jvm,thread_pool
cluster = True

When one of instance ES doesn't response by timeout then diamond doesn't send stats from all other instances too. It confused! Now when one of process has a problem it looks like as whole cluster are broken.

[shortdudey123]

Do you have any logs that show where in the collector the timeout happened?
I am guessing that there is an uncaught exception that causes the entire collector to bomb out if a timeout happens